Arkandi, Indi (2015) Analisis Faktor Risiko Kematian Ibu Dan Kematian Bayi Dengan Pendekatan Regresi Poisson Bivariat Di Provinsi Jawa Timur Tahun 2013. Undergraduate thesis, Institut Technology Sepuluh Nopember.
Preview |
Text
1311100037-Undergraduate Thesis.pdf - Published Version Download (2MB) | Preview |
Abstract
Angka kematian ibu dan bayi merupakan salah satu indikator yang paling
menonjol untuk menilai derajat kesehatan masyarakat. Tingginya angka
kematian ibu dan kematian bayi di Indonesia salah satunya berasal dari
Provinsi Jawa Timur. Perlu adanya tindakan dari pemerintah untuk menekan
angka kematian ibu dan bayi di Jawa Timur. Kematian ibu dan kematian
bayi merupakan dua hal yang saling berkaitan karena selama masa
kandungan, gizi yang diperoleh janin disalurkan dari tubuh ibu melalui
plasenta sehingga kondisi ibu selama masa kehamilan akan berpengaruh
pada janin dan bayi yang akan dilahirkannya. Sehingga perlu dilakukan
penelitian untuk menganalisis faktor-faktor yang mempengaruhi kedua
angka kematian tersebut dengan metode Regresi Poisson Bivariat menggunakan
algoritma Expectation Maximization. Terdapat tiga buah model
dengan nilai kovarians yang berbeda pada Regresi Poisson Bivariat. Oleh
karena itu perlu memilih salah satu dari ketiga model tersebut. Setelah
mendapat model terbaik dengan kriteria AIC, diketahui bahwa model
terbaik adalah model dengan nilai kovarians antara jumlah kematian ibu
dan bayi adalah fungs variabel bebas. Pada model terbaik variabel yang
signifikan mempengaruhi jumlah kematian ibu adalah persentase kunjungan
ibu hamil dengan K4 dan persentase ibu hamil yang mendapat
tablet Fe3. Sedangkan variabel yang signifikan mempengaruhi jumlah
kematian bayi adalah semua variabel prediktor kecuali variabel persentase
kunjungan ibu hamil dengan K1.
====================================================================================================
Maternal and infant mortality rate is one of the most prominent indicators
to assess the degree of public health. East Java is one of the provinces that
contribute great number of maternal and infants’ mortality in Indonesia.
The government needs to reduce the number of maternal and infant mortality
in East Java. Maternal and infants’ mortality are interrelated one another,
because during gestation period, the fetus obtains the nutrition through
the placenta from the mother's body, so that the condition of the mother
during pregnancy will affect the fetus and the baby. Therefore, it is necessary
to analyze the factors that affect both mortalities simultaneously with
Bivariate Poisson Regression method, by using Expectation Maximization
algorithm. There are three models with different covarance values on
Bivariate Poisson Regression. Therefore, it is necessary to choose one of
three models. After obtaining the best model with AIC criteria, it is known
that the best model to be used is the model, which is using covariates on
covariance. In the best model, variables that significantly affect the number
of maternal mortality is the visit percentage of pregnant women with K4
and the percentage of pregnant women who received Fe3 tablets. While variables
that signifycantly affect the number of infants’ mortality is the visit
percentage of pregnant women with K4, the percentage of pregnant women
who receive Fe3 tablets, the percentage of obstetrical complications, the
percentage of births attended by skilled health personnel, the percentage of
active participation in family planning, as well as the percentage of clean
and healthy behavior in household.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSt 519.536 Ark a |
| Uncontrolled Keywords: | Algoritma EM, Jawa Timur, Kematian Bayi, Kematian Ibu, Regresi Poisson Bivariat |
| Subjects: | Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression |
| Divisions: | Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Mr. Tondo Indra Nyata |
| Date Deposited: | 21 Nov 2018 03:49 |
| Last Modified: | 21 Nov 2018 03:49 |
| URI: | http://repository.its.ac.id/id/eprint/59976 |
Actions (login required)
 |
View Item |